Schizophrenia Research 117 (2010) 92–98

Contents lists available at ScienceDirect

Schizophrenia Research

j ourna l homepage: www.e lsev ie r.com/ locate /schres

Abnormal glycosylation of EAAT1 and EAAT2 in prefrontal cortex of elderlypatients with schizophrenia

Deborah Bauer a,b,⁎, Vahram Haroutunian c,James H. Meador-Woodruff a,b, Robert E. McCullumsmith a

a Department of Psychiatry and Behavioral Neurobiology, University of Alabama at Birmingham, Birmingham, AL, United Statesb Program in Neuroscience, University of Michigan, Ann Arbor, MI, United Statesc Department of Psychiatry, Mount Sinai School of Medicine, New York, NY, United States

a r t i c l e i n f o

⁎ Corresponding author. Department of Psychiatry,at Birmingham, SC560, 1530 3rd Ave S., Birmingham, AStates. Fax: +1 205 975 4879.

E-mail address: debbauer@uab.edu (D. Bauer).

0920-9964/$ – see front matter. Published by Elseviedoi:10.1016/j.schres.2009.07.025

a b s t r a c t

Article history:Received 25 June 2009Received in revised form 29 July 2009Accepted 31 July 2009Available online 27 August 2009

The excitatory amino acid transporters (EAATs) are a family of molecules that are essential forregulation of synaptic glutamate levels. The EAATs may also be regulated by N-glycosylation, aposttranslational modification that is critical for many cellular functions including localizationin the plasma membrane. We hypothesized that glycosylation of the EAATs is abnormal inschizophrenia. To test this hypothesis, we treated postmortem tissue from the dorsolateralprefrontal and anterior cingulate cortices of patients with schizophrenia and comparisonsubjects with deglycosylating enzymes. We then measured the resulting shifts in molecularweight of the EAATs usingWestern blot analysis to determine the mass of glycans cleaved fromthe transporter. We found evidence for less glycosylation of both EAAT1 and EAAT2 inschizophrenia. We did not detect N-linked glycosylation of EAAT3 in either schizophrenia orthe comparison subjects in these regions. Our data suggest an abnormality of posttranslationalmodification of glutamate transporters in schizophrenia that suggests a decreased capacity forglutamate reuptake.

Published by Elsevier B.V.

Keywords:GLASTGLT-1EAAC1DeglycosylationAnterior cingulate cortexDorsolateral prefrontal cortex

1. Introduction

Glycosylation of proteins is a posttranslational modificationthat plays a role in molecular trafficking, protein folding,endocytosis, receptor activation, signal transduction, and celladhesion (Ohtsubo and Marth, 2006). Abnormalities of glyco-sylation can lead to a number of cellular storage disordersincluding Gaucher's, Niemann–Pick type C, Sandhoff's, and Tay–Sach's diseases, as well as other congenital disorders ofglycosylation (Ohtsubo and Marth, 2006). Disruptions in glyco-sylation have also been implicated in Alzheimer's disease(Takeuchi and Yamagishi, 2009), Huntington's disease (Hunget al., 1980), and schizophrenia (Narayan et al., 2008).

University of AlabamaL 35294-0017, United

r B.V.

Two common forms of protein glycosylation include N-linkedglycosylation and O-linked glycosylation. N-linked glycosyl-ation is the covalent linkage of oligosaccharides to asparagineresidues of proteins. N-glycosyl residues are processed asproteins are trafficked through the endoplasmic reticulumand golgi. The excitatory amino acid transporters (EAATs) areN-glycosylated proteins that transport extracellular gluta-mate out of the synapse and thus are critical for glutamatergicsignaling. However, glycosylation of the EAATs in humanbrain has not been evaluated.

EAAT1 is variably expressed throughout the cortex inastroglia (Rothstein et al., 1994; Chaudhry et al., 1995; Kondoet al., 1995; Lehre et al., 1995; Gegelashvili et al., 1996; Schmittet al., 1997; Williams et al., 2005). GLAST, the rodent form ofEAAT1, exists as two isoforms, 70-kDa and 64-kDa, which differonly by the degree of N-glycosylation at Asn206 and Asn195(Conradt et al., 1995; Schulte and Stoffel, 1995). Glycosylation ofthis transporter may serve an important functional role becausenonglycosylated GLAST does not form homomultimers, which

Table 1Subject characteristics.

Comparison group Schizophrenia

Region ACC DLPFC ACC DLPFCN 34 32 34 33Sex 14 m/20 f 12 m/20 f 24 m/10 f 23 m/10 fTissue pH 6.4±0.2 6.5±0.2 6.4±0.3 6.4±0.3PMI (h) 8.3±6.7 8.2±6.8 13.4±8.1 12.5±6.7Age (years) 78±14 78±14 74±12 74±12On/off Rx 0/34 0/32 23/11 22/11

Values presented as mean±standard deviation.Abbreviations: anterior cingulate cortex (ACC), dorsolateral prefrontal cortex(DLPFC), male (m), female (f), antipsychotic medication (Rx), postmorteminterval (PMI).

93D. Bauer et al. / Schizophrenia Research 117 (2010) 92–98

are the native conformation of GLAST in vivo (Conradt et al.,1995). In addition, glycosylation of GLAST has been correlatedwith trafficking of GLAST to plasma membrane and increasedglutamate uptake (Escartin et al., 2006).

EAAT2 is an astrocytic transporter responsible for themajority of glutamate uptake in the cortex. Deglycosylation ofthe rodent isoforms of EAAT2 (GLT-1) resulted in a ~10–15 kDashift in molecular weight of the monomer band (Kalandadzeet al., 2004). There is a conflicting literature describing thefunctional effects of EAAT2 glycosylation. One group found thatglycosylation-deficient GLT-1 had a decreased rate of glutamatetransport due to decreased expression in the plasmamembrane(Trotti et al., 2001). This may be attributed to retention of GLT-1in the endoplasmic reticulum, becausemutant GLT-1 expressingan altered extracellular leucine-based motif is immaturelyglycosylated and retained in the ER (Kalandadze et al., 2004).However, another group found no effect of N-glycosylation onthe trafficking or transport activity of GLT-1 in transfected BHKcells, but increased stability at theplasmamembrane,whichmaybe critical for transporter localization in vivo (Raunser et al.,2005).

EAAT3 is a neuronal glutamate transporter expressed in thecortex. In rat C-6 glioma cells, EAAC1 (the rodent form ofEAAT3) is N-glycosylated with high mannose-containing side-chains and processed into complex chains, coinciding withinsertion into the plasmamembrane (Yang and Kilberg, 2002).A shift of approximately 5 kDa was detected when EAAT3immunoprecipitated from human brain synaptosomes wastreated with Endoglycosidase F (Shashidharan et al., 1997).

We previously reported alterations in EAAT1 and EAAT3protein in prefrontal cortex in schizophrenia, suggestingdiminished EAAT-mediated glutamate reuptake as a part ofthe pathophysiology of this illness (Bauer et al., 2008).However, localization of the transportersmay be as importantas overall protein levels. Altered EAAT localizationmay lead toglutamate spillover into the extrasynaptic space and adjacentsynapses, causing loss of input specificity (Overstreet et al.,1999; Tsvetkov et al., 2004; Marcaggi and Attwell, 2007).Since glycosylation is important for targeting of the EAATs tothe plasma membrane, abnormal glycosylation of theseproteins may play a role in schizophrenia.

Glycobiology is a growing field with an increasing numberof tools. The enzyme peptide-N4-(N-acetyl-beta-glucosami-nyl) asparagine amidase F (PNGase F) cleaves N-linked sugarsoff of proteins attached at asparagine residues. Endoglycosi-dase H (EndoH) cleaves hybrid and highmannose-containingresidues from glycoproteins, and is therefore specific toimmaturely glycosylated proteins that have not been pro-cessed beyond the endoplasmic reticulum. The removal ofglycans is often substantial enough to detect a change inmolecular weight of proteins when measured by Westernblot analysis. In this study, we assessed glycosylation ofEAAT1, EAAT2, and EAAT3 through enzymatic deglycosyla-tion in schizophrenia and a comparison group.

2. Materials and methods

2.1. Subjects

Subjects from theMount Sinai Medical Center Schizophre-nia Brain Bank were studied (Table 1), including 35 indivi-

duals diagnosed with schizophrenia and 33 comparisonsubjects. Subjects were diagnosed with schizophrenia if thepresence of schizophrenic symptoms was documented beforeage 40, the medical records contained evidence of psychoticsymptoms and at least 10 years of psychiatric hospitalizationwith diagnosis of schizophrenia, and a DSM-III-R diagnosis ofschizophreniawas agreeduponby twoexperienced clinicians.Diagnostic groups did not significantly differ for age, sex,postmortem interval, and tissue pH. Upon neuropathologicalexamination, no evidence of Alzheimer or other neurodegen-erative diseasewas found. The brain banking procedureswereapproved by the Mount Sinai School of Medicine InstitutionalReview Board.

2.2. Tissue preparation

Brains were obtained after autopsy and one hemispherewas cut coronally into ~0.8–1 cm3 slabs and flash frozen. Graymatter was dissected from anterior cingulate cortex (ACC)(n=68) and dorsolateral prefrontal cortex (DLPFC) (n=66).ACC was dissected at the level of the genu of the corpuscallosum. Tissue blockswere dissected from the dorsal surfaceof the corpus callosum extending 12–15 mm dorsally andextending 12–15 mm laterally from the midline. DLPFC wasdissected corresponding to Brodmann area 46 and measuring≈1.5 cm along the cortical surface as described by Rajkowskaand Goldman-Rakic (Rajkowska and Goldman-Rakic, 1995).Approximately 1 cm3 of frozen tissue was pulverized in liquidnitrogen, then homogenized (10% wt/vol) in 5 mM Tris–HCl(pH 7.4) with 320 mM sucrose and 1 protease inhibitor tablet(Complete mini, Roche Diagnostics, Manheim, Germany) per10ml for 30 s with a polytron homogenizer (Fisher Scientific,Pittsburgh, Pennsylvania) and stored at −80 °C in 0.5mlaliquots. To determine protein concentrations, assay by theBradford method (Bradford, 1976) was performed on thesehomogenates.

2.3. Deglycosylation

16 μg of protein for each sample was added to 6.7 μl 5×reaction buffer (QA Bio), 1.7 μl denaturation solution (2% SDS/1 M β mercaptoethanol) (QA Bio), and adjusted to volumewith deionized water. Samples were then incubated at 70 °Cfor 10 min. Samples were cooled to room temperature andincubated with 1.3 μl Endoglycosidase H or 1.3 μl PNGase F

94 D. Bauer et al. / Schizophrenia Research 117 (2010) 92–98

and 1.7 μl 15% triton X-100 (QA Bio) at 37 °C for 12 h. Non-enzyme-treated samples were prepared identically to theenzyme-treated samples with the same buffers except thatthey were incubated with water instead of the deglycosylat-ing enzymes.

2.4. Electrophoresis

NuPAGE sample reducing agent (Invitrogen), and NuPAGELDS sample buffer (Invitrogen) were added to the samples,which were then incubated at 70 °C for 10 min. The NovexMini Cell NuPAGE system (Invitrogen) with 4–12% Bis–Trisgradient polyacrylamide gels (Invitrogen) was used and 8 μgof protein was added per lane. A molecular mass standardwas run on each gel (Kaleidescope prestained standards, Bio-Rad). Gels were suspended in a bath of NuPAGE MES SDSrunning buffer (Invitrogen) with 500 μl NuPAGE antioxidant(Invitrogen) during electrophoresis.

2.5. Western blot analysis

Following electrophoresis, proteins were transferred ontoImmobilon-FL PVDF membranes (Millipore) using a semi-drytransfer apparatus (Bio-Rad). After electroblot transfer,membranes were washed twice and incubated with Li-CorBlocking Buffer (Li-Cor Biosciences) for 1 h at room temper-ature with rocking to block nonspecific antibody binding.Membranes were incubated with either a rabbit polyclonalantibody to EAAT1 (Santa Cruz sc-15316) diluted 1:1000,rabbit polyclonal antibody to EAAT2 (Santa Cruz sc-15317)diluted 1:1000, mouse monoclonal antibody to EAAT3(Chemicon MAB1578) diluted 1:1000, rabbit polyclonalantibody to EAAT3 (Santa Cruz sc-25658) diluted 1:500, orrabbit polyclonal antibody to EAAT3 (Alpha Diagnostics#EAAC11-A) diluted 1:500 in blocking buffer with 0.1%Tween overnight at 4 °C with rocking. Next, the membraneswere washed three times for 10 min in tris-buffered salinewith 0.1% Tween (TBST), then rocked for 15 min at roomtemperature with anti-rabbit or anti-mouse IR-Dye 800CWsecondary antibody (Li-Cor Biosciences) diluted 1:10,000 inblocking buffer with 0.1% Tween. Membranes were washedthree times for 10 min in TBST then washed 5 times indeionized water and allowed to dry for 3–5 min beforescanning (infrared imaging system; Li-Cor Biosciences).

2.6. Data analysis

Membranes probed with infrared-labeled secondary anti-bodies were scanned using a Li-Cor Odyssey scanner, and themigration distance for each protein band was measured inpixels using the Odyssey 2.1 software package. Migrationdistance was converted to molecular mass by plotting therelativemigration of themolecularmass standards against thelog of their molecular masses, and fitting the relativemigration of the bands of interest to that standard curve(Jarvie et al., 1988). Band shift was measured as molecularmass of the control band minus the molecular mass of theenzyme-treated band in the adjacent lane, as describedpreviously (Jimenez-Huete et al., 1998; Nielsen et al., 2004;Toledo et al., 2005). EAAT1 and EAAT2 migrate as bothmonomers and multimers (Bauer et al., 2008), and the

molecular mass shifts of monomers and multimers were ana-lyzed separately.

2.7. Statistical analysis

All statistical analyses were performed using Statistica(StatSoft, Tulsa, Oklahoma). Outliers more than 6 standarddeviations from themeanwere excluded. Correlation analysiswas performed to determine associations between thedependent variable, molecular mass shift and age, PMI, andpH. We analyzed deglycosylation induced changes in molec-ular mass using analysis of variance (ANOVA), or analysis ofcovariance (ANCOVA) when significant correlations weredetected. To test for possible medication effects, patients withschizophrenia off antipsychotic medication for at least6 weeks prior to death were compared to patients on anti-psychotic medication within 6 weeks of death.

3. Results

When samples were treated with EndoH, none of thetransporters exhibited shifts in molecular mass of eithermonomeric or multimeric forms (Fig. 1). When samples weretreatedwith PNGase F, EAAT1 and EAAT2 exhibited detectableshifts in molecular mass for both monomeric and multimericforms (Fig. 1). However, EAAT3 did not shift when treatedwith PNGase F (Fig. 1). Because we were concerned that thelack of shift could be due to a loss of an epitope followingenzymatic digestion, we performed Western blots with twoadditional EAAT3 antibodies raised against different epitopes,and did not detect shifts in EAAT3 (data not shown).

We examined the effects of PNGase treatment onmonomer and multimer forms of EAAT1 and EAAT2 inschizophrenia and comparison subjects. There were generallyno correlations detected between age, PMI (which differsbetween diagnosis groups (F(1, 66)=7.3766, p=0.00843)),or pH and our dependentmeasureswith the exception of shiftof EAAT1 monomer in the DLPFC: (R=0.28, pb0.05). Wefound less of a molecular mass shift for the EAAT1 monomerin schizophrenia in the ACC (F(1, 61)=6.40; pb0.05) (Fig. 2).We found less of a mass shift for the EAAT2 multimer inschizophrenia in the DLPFC (F(1,52)=9.41; pb0.05) (Fig. 2).There was no effect of medication status in the subjects withschizophrenia on either of these dependent measures (EAAT1monomer in ACC (F(1, 29)=0.06, p=0.80), EAAT2 multimerin DLPFC (F(1, 27)=1.04, p=0.32)) (Fig. 3).

4. Discussion

Previous work has demonstrated altered glycosylation ofseveral proteins in schizophrenia. Increases in theplasmaactivityof the glycosylating enzyme alpha 2,6 sialyltransferase and inserum levels of alpha 2 and beta globulins have been found inschizophrenia (Varma and Hoshino, 1980; Maguire et al., 1997).In addition, a decrease in the number of cells expressingpolysialatedneural cell adhesionmolecule (NCAM)wasdetectedin the hilus of the hippocampus without an overall change inNCAM expression (Barbeau et al., 1995). We detected twoadditional proteins that have altered glycosylation in schizo-phrenia, suggesting that deficits in glycosylationmay have a rolein the pathophysiology of this illness.

Fig. 1. Western blots of deglycosylated EAATs. A. Western blot analysis of EAAT1, EAAT2, and EAAT3 deglycosylated with the enzymes Endoglycosidase H andPNGase F. EndoH and PNGase F lanes indicate enzyme-treated samples. EndoH control and PNGase F control lanes were treated identically to the correspondingenzyme-treated samples except the enzymes were omitted. Molecular masses of EAAT1 and EAAT2monomers andmultimers were shifted in the PNGase F treatedlanes. No shift was detected for EAAT3. B. Western blot analysis of EAAT1 deglycosylated with EndoH and PNGase F in anterior cingulate cortex and dorsolateralprefrontal cortex from a patient with schizophrenia and a comparison subject. Abbreviations: kDa (kilodaltons), excitatory amino acid transporter (EAAT),endoglycosidase H (EndoH), peptide-N4-(N-acetyl-beta-glucosaminyl) asparagine amidase F (PNGase F).

95D. Bauer et al. / Schizophrenia Research 117 (2010) 92–98

We found that EAAT1 and EAAT2, but not EAAT3, are N-glycosylated in the humanbrain. Given thatwe found changes inthe glial (EAAT1 and EAAT2) but not neuronal (EAAT3)transporters, it is possible that themechanisms for glycosylationdeficits in schizophrenia are glia-specific. Our EAAT3 finding issurprising, given that in the rodent EAAT3 (EAAC1) is glycosy-lated (Yang and Kilberg, 2002), and that another studydemonstrated deglycosylation of EAAT3 in an immunoprecipi-

tated fraction from human brain synaptosomes (Shashidharanet al., 1997). These divergentfindingsmight be due to the type ofdeglycosylating enzyme used, or that immunoprecipitatingEAAT3 from synaptosomes significantly enriched EAAT3, allow-ing detection of subtle changes that might not be apparent intissue homogenate. It may be that a small subset of EAAT3 isglycosylated in human brain, while the majority of EAAT3 isunglycosylated, and thus not detectable with our approach.

Fig. 2.Molecular mass shifts of EAAT1 and EAAT2 in schizophrenia and a comparison group following enzymatic deglycosylation with PNGase F. Data expressed asmeans+/−standard error of the mean. Asterisks indicate a significant difference between schizophrenia and comparison subjects (pb0.05). Abbreviations: kDa(kilodaltons), anterior cingulate cortex (ACC), dorsolateral prefrontal cortex (DLPFC), excitatory amino acid transporter (EAAT), endoglycosidase H (EndoH),peptide-N4-(N-acetyl-beta-glucosaminyl) asparagine amidase F (PNGase F).

96 D. Bauer et al. / Schizophrenia Research 117 (2010) 92–98

Alternatively, the absence of a molecular mass shift in EAAT3could be due to a loss of epitopes associated with glycosylationfor the EAAT3 antibodies following enzymatic digestion(Levenson et al., 2002; Holmseth et al., 2005). For example,an antibody might bind to a glycosylated residue of EAAT3 andlose antigenicity if that glycan is cleaved. However, we feel thisis unlikely because we did not detect a shift using any of thethree EAAT3 antibodies raised to different epitopes and one ofthese epitopes does not contain any putative N-glycosylationsites.

We found that EAAT1 has fewer sugar residues added by N-glycosylation in schizophrenia. The change in molecular massshift of EAAT1 was relatively small (~5%) between schizophre-nia and the comparison group. It is difficult to determine if sucha small change in glycosylation is of physiological significance.However, preclinical data suggest that a small change inglycosylation can have a strong effect on glutamate uptake.For example, activation of astrocytes with ciliary neurotrophicfactor (CNTF) results in a small increase in glycosylation ofGLAST (~8% shift), with increased localization of GLAST to lipidrafts at the cell surface and increased glutamate reuptake,resulting in a 67% decrease in extracellular glutamate levelsupon quinolinate evoked glutamate release (Escartin et al.,2006). This suggests that thedecrease in glycosylationof EAAT1that we detectedmay significantly impact glutamate reuptake.

We also found evidence for less glycosylation of the otherastrocytic transporter, EAAT2, in schizophrenia. The differencein molecular mass shift between groups was larger for EAAT2(~13%) than for EAAT1, although the functional effect of thislarger shift is not known. Less glycosylation of EAAT2 mightreflect decreased glutamate reuptake, since altered glycosyla-tion of EAAT2 is associated with ER retention and decreasedplasma membrane expression, and trafficking of EAAT2 to theplasmamembrane is necessary for EAAT2 mediated glutamatereuptake (Trotti et al., 2001; Kalandadze et al., 2004).

One potential mechanism for the decreases in glycosyla-tion could be altered splice variant expression. EAAT1 andEAAT2 can both be alternatively spliced to skip exon 9, whichcontains an ER exit motif. These splice variants are retained inthe ER, and cause ER retention of any full length variants withwhich they dimerize (Kalandadze et al., 2004). The transpor-ters that are retained in the ER are less glycosylated thantransporters that are not retained (Kalandadze et al., 2004).Thus, it is possible that the decreases in glycosylation that wefound are due to increased expression of these exon skippingvariants. In fact, we found increases in the exon 9 skippingvariant of EAAT2, which could explain the decrease we foundin EAAT2 glycosylation (unpublished observation).

It is also possible that the changes in glycosylationwe foundare due to changes in the levels or activity of the glycosyl

Fig. 3.Molecular mass shifts of EAAT1 and EAAT2 following enzymatic deglycosylation with PNGase F in patients with schizophrenia off or on medication 6 weeksprior to death. Data expressed as means+/−standard error of themean. Abbreviations: kDa (kilodaltons), anterior cingulate cortex (ACC), dorsolateral prefrontalcortex (DLPFC), excitatory amino acid transporter (EAAT).

97D. Bauer et al. / Schizophrenia Research 117 (2010) 92–98

transferases that attach glycans to the proteins. Few studieshave investigated glycosyl transferases in schizophrenia. Anincrease has been detected in the plasma activity of theglycosylating enzyme alpha 2,6 sialyltransferase (Maguireet al., 1997). An increase in activity of a glycosyl transferase isunlikely to explain a decrease in glycosylation, but it is possiblethat other glycosyl transferases are decreased in schizophrenia.

Since most of the patients with schizophrenia were treatedwith antipsychotic medications, the reductions we found inglycosylation could be due to amedication effect. However, wedid not find any effects of medication on molecular mass shiftwhen comparing patients onmedication 6 weeks prior to deathto patients off medication at least 6 weeks prior to death.

The reductions in EAAT1 and EAAT2 glycosylation suggestdecreased plasmamembrane expression of these transporters.Altered localization of EAAT1 and EAAT2, combined with thedecreased EAAT1 protein expression we previously described(Bauer et al., 2008), suggests that there is decreased perisy-naptic glutamate reuptake into astrocytes in schizophrenia.The glutamate transporters are important for maintaining lowsynaptic glutamate levels by buffering and transporting sy-naptic glutamate (Tong and Jahr, 1994; Tzingounis andWadiche, 2007). Diminished perisynaptic reuptake and buff-ering may lead to glutamate spillover and loss of input spe-cificity (Overstreet et al., 1999; Tsvetkov et al., 2004). Our datasuggesting decreased glutamate reuptake support a hypothesisof increased synaptic glutamate levels and/or glutamatespillover in schizophrenia. Consistent with this hypothesis,EAAT1 deficient mice exhibit endophenotypes including self-neglect, social withdrawal, and impaired learning, suggestingthat schizophrenia-associated rodent endophenotypes can bemodeled by disruption of EAAT1-mediated glutamate reuptake(Karlsson et al., 2009). This hypothesis is further supported by areport of a subject with schizophrenia who has a partial dele-tion of the EAAT1 gene (Walsh et al., 2008). Finally, our datasuggest that reducing synaptic glutamate could be a usefulstrategy in the treatment of schizophrenia. One study using anmGluR2/3 agonist, which decreases glutamate release, hadantipsychotic effects in schizophrenia (Patil et al., 2007). Takentogether, these data support a role for diminished glutamatereuptake in the pathophysiology of schizophrenia.

Role of funding sourceFunding for this study was provided by the NIMH grants MH53327,

MH78378, and MH074016. The NIMH had no further role in study design; inthe collection, analysis and interpretation of data; in thewriting of the report;and in the decision to submit the paper for publication.

ContributorsDrs. McCullumsmith and Meador-Woodruff helped design the study and

provided intellectual contributions. Dr. Harotunian provided the tissue for thestudy.Ms. Bauerhelpeddesign the study,performed thedeglycosylationassays,performed the statistical analyses andwrote thefirst draft of themanuscript. Allauthors contributed to and have approved the final manuscript.

Conflict of interestAll authors declare they have no conflicts of interest.

AcknowledgmentsThis work is supported byMH53327 (JMW), MH78378 (DEB), MH064673

and MH066392 (VH) and MH074016 (REM).

References

Barbeau, D., Liang, J.J., Robitalille, Y., Quirion, R., Srivastava, L.K., 1995. Decreasedexpression of the embryonic form of the neural cell adhesion molecule inschizophrenic brains. Proc. Natl. Acad. Sci. U. S. A. 92 (7), 2785–2789.

Bauer, D., Gupta, D., Harotunian, V., Meador-Woodruff, J.H., McCullumsmith,R.E., 2008. Abnormal expression of glutamate transporter and trans-porter interacting molecules in prefrontal cortex in elderly patients withschizophrenia. Schizophr. Res. 104 (1–3), 108–120.

Bradford, M.M., 1976. A rapid and sensitive method for the quantitation ofmicrogram quantities of protein utilizing the principle of protein–dyebinding. Anal. Biochem. 72, 248–254.

Chaudhry, F.A., Lehre, K.P., van Lookeren Campagne, M., Ottersen, O.P., Danbolt,N.C., Storm-Mathisen, J., 1995. Glutamate transporters in glial plasmamembranes: highly differentiated localizations revealed by quantitativeultrastructural immunocytochemistry. Neuron 15 (3), 711–720.

Conradt,M., Storck, T., Stoffel,W., 1995. Localization ofN-glycosylation sites andfunctional role of the carbohydrate units of GLAST-1, a cloned rat brain L-glutamate/L-aspartate transporter. Eur. J. Biochem. 229 (3), 682–687.

Escartin, C., Brouillet, E., Gubellini, P., Trioulier, Y., Jacquard, C., Smadja, C.,Knott, G.W., Kerkerian-Le Goff, L., Deglon, N., Hantraye, P., Bonvento, G.,2006. Ciliary neurotrophic factor activates astrocytes, redistributes theirglutamate transporters GLAST and GLT-1 to raft microdomains, andimproves glutamate handling in vivo. J. Neurosci. 26 (22), 5978–5989.

Gegelashvili, G., Civenni, G., Racagni, G., Danbolt, N.C., Schousboe, I., Schousboe,A., 1996. Glutamate receptor agonists up-regulate glutamate transporterGLAST in astrocytes. NeuroReport 8 (1), 261–265.

Holmseth, S., Dehnes, Y., Bjornsen, L.P., Boulland, J.L., Furness, D.N., Bergles, D.,Danbolt, N.C., 2005. Specificity of antibodies: unexpected cross-reactivity ofantibodies directed against the excitatory amino acid transporter 3 (EAAT3).Neuroscience 136 (3), 649–660.

98 D. Bauer et al. / Schizophrenia Research 117 (2010) 92–98

Hung, W.Y., Mold, D.E., Tourian, A., 1980. Huntington's-chorea fibroblasts.Cellular protein glycosylation. Biochem. J. 190 (3), 711–719.

Jarvie, K.R., Niznik, H.B., Seeman, P., 1988. Dopamine D2 receptor bindingsubunits of Mr congruent to 140, 000 and 94, 000 in brain: deglycosyla-tion yields a common unit of Mr congruent to 44, 000. Mol. Pharmacol.34 (2), 91–97.

Jimenez-Huete, A., Lievens, P.M., Vidal, R., Piccardo, P., Ghetti, B., Tagliavini, F.,Frangione, B., Prelli, F., 1998. Endogenous proteolytic cleavage of normaland disease-associated isoforms of the human prion protein in neuraland non-neural tissues. Am. J. Pathol. 153 (5), 1561–1572.

Kalandadze, A., Wu, Y., Fournier, K., Robinson, M.B., 2004. Identification ofmotifs involved in endoplasmic reticulum retention-forward traffickingof the GLT-1 subtype of glutamate transporter. J. Neurosci. 24 (22),5183–5192.

Karlsson, R.M., Tanaka, K., Saksida, L.M., Bussey, T.J., Heilig, M., Holmes, A., 2009.Assessment of glutamate transporter GLAST (EAAT1)-deficient mice forphenotypes relevant to the negative and executive/cognitive symptoms ofschizophrenia. Neuropsychopharmacology 34 (6), 1578–1589.

Kondo, K., Hashimoto, H., Kitanaka, J., Sawada, M., Suzumura, A., Marunouchi,T., Baba, A., 1995. Expression of glutamate transporters in cultured glialcells. Neurosci. Lett. 188 (2), 140–142.

Lehre, K.P., Levy, L.M., Ottersen, O.P., Storm-Mathisen, J., Danbolt, N.C., 1995.Differential expression of two glial glutamate transporters in the rat brain:quantitative and immunocytochemical observations. J. Neurosci. 15 (3 Pt 1),1835–1853.

Levenson, J., Weeber, E., Selcher, J.C., Kategaya, L.S., Sweatt, J.D., Eskin, A.,2002. Long-term potentiation and contextual fear conditioning increaseneuronal glutamate uptake. Nat. Neurosci. 5 (2), 155–161.

Maguire, T.M., Thakore, J., Dinan, T.G., Hopwood, S., Breen, K.C., 1997. Plasmasialyltransferase levels in psychiatric disorders as a possible indicator ofHPA axis function. Biol. Psychiatry 41 (11), 1131–1136.

Marcaggi, P., Attwell, D., 2007. Short- and long-term depression of ratcerebellar parallel fibre synaptic transmission mediated by synapticcrosstalk. J. Physiol. 578 (Pt 2), 545–550.

Narayan, S., Head, S.R., Gilmartin, T.J., Dean, B., Thomas, E.A., 2008. Evidence fordisruption of sphingolipid metabolism in schizophrenia. J. Neurosci. Res.

Nielsen, D., Gyllberg, H., Ostlund, P., Bergman, T., Bedecs, K., 2004. Increasedlevels of insulin and insulin-like growth factor-1 hybrid receptors anddecreased glycosylation of the insulin receptor alpha- and beta-subunits inscrapie-infected neuroblastoma N2a cells. Biochem. J. 380 (Pt 2), 571–579.

Ohtsubo, K., Marth, J.D., 2006. Glycosylation in cellular mechanisms of healthand disease. Cell 126 (5), 855–867.

Overstreet, L.S., Kinney, G.A., Liu, Y.B., Billups, D., Slater, N.T., 1999. Glutamatetransporters contribute to the time course of synaptic transmission incerebellar granule cells. J. Neurosci. 19 (21), 9663–9673.

Patil, S.T., Zhang, L., Martenyi, F., Lowe, S.L., Jackson, K.A., Andreev, B.V.,Avedisova, A.S., Bardenstein, L.M., Gurovich, I.Y., Morozova, M.A.,Mosolov, S.N., Neznanov, N.G., Reznik, A.M., Smulevich, A.B., Tochilov,V.A., Johnson, B.G., Monn, J.A., Schoepp, D.D., 2007. Activation of mGlu2/3receptors as a new approach to treat schizophrenia: a randomized Phase2 clinical trial. Nat. Med. 13 (9), 1102–1107.

Rajkowska, G., Goldman-Rakic, P.S., 1995. Cytoarchitectonic definition ofprefrontal areas in the normal human cortex: I. Remapping of areas 9 and46 using quantitative criteria. Cereb. Cortex 5 (4), 307–322.

Raunser, S., Haase, W., Bostina, M., Parcej, D.N., Kuhlbrandt, W., 2005. High-yield expression, reconstitution and structure of the recombinant, fully

functional glutamate transporter GLT-1 from Rattus norvegicus. J. Mol.Biol. 351 (3), 598–613.

Rothstein, J.D., Martin, L., Levey, A.I., Dykes-Hoberg, M., Jin, L., Wu, D., Nash,N., Kuncl, R.W., 1994. Localization of neuronal and glial glutamatetransporters. Neuron 13 (3), 713–725.

Schmitt, A., Asan, E., Puschel, B., Kugler, P., 1997. Cellular and regionaldistribution of the glutamate transporter GLAST in the CNS of rats:nonradioactive in situ hybridization and comparative immunocytochem-istry. J. Neurosci. 17 (1), 1–10.

Schulte, S., Stoffel, W., 1995. UDP galactose:ceramide galactosyltransferase andglutamate/aspartate transporter. Copurification, separation and character-ization of the two glycoproteins. Eur. J. Biochem. 233 (3), 947–953.

Shashidharan, P., Huntley, G.W., Murray, J.M., Buku, A., Moran, T., Walsh, M.J.,Morrison, J.H., Plaitakis, A., 1997. Immunohistochemical localization ofthe neuron-specific glutamate transporter EAAC1 (EAAT3) in rat brainand spinal cord revealed by a novel monoclonal antibody. Brain Res.773 (1–2), 139–148.

Takeuchi, M., Yamagishi, S., 2009. Involvement of toxic AGEs (TAGE) in thepathogenesis of diabetic vascular complications and Alzheimer's disease.J. Alzheimer's Dis. 16 (4), 845–858.

Toledo, J.R., Sanchez, O., Montesino Segui, R., Fernandez Garcia, Y., Rodriguez,M.P., Cremata, J.A., 2005. Differential in vitro and in vivo glycosylation ofhuman erythropoietin expressed in adenovirally transduced mousemammary epithelial cells. Biochim. Biophys. Acta 1726 (1), 48–56.

Tong, G., Jahr, C.E., 1994. Block of glutamate transporters potentiatespostsynaptic excitation. Neuron 13 (5), 1195–1203.

Trotti, D., Aoki, M., Pasinelli, P., Berger, U.V., Danbolt, N.C., Brown Jr., R.H.,Hediger, M.A., 2001. Amyotrophic lateral sclerosis-linked glutamatetransporter mutant has impaired glutamate clearance capacity. J. Biol.Chem. 276 (1), 576–582.

Tsvetkov, E., Shin, R.M., Bolshakov, V.Y., 2004. Glutamate uptake determinespathway specificity of long-term potentiation in the neural circuitry offear conditioning. Neuron 41 (1), 139–151.

Tzingounis, A.V., Wadiche, J.I., 2007. Glutamate transporters: confiningrunaway excitation by shaping synaptic transmission. Nat. Rev. Neurosci.8 (12), 935–947.

Varma, R., Hoshino, A.Y., 1980. Serum glycoproteins in schizophrenia.Carbohydr. Res. 82 (2), 343–351.

Walsh, T., McClellan, J.M., McCarthy, S.E., Addington, A.M., Pierce, S.B., Cooper,G.M., Nord, A.S., Kusenda, M., Malhotra, D., Bhandari, A., Stray, S.M.,Rippey, C.F., Roccanova, P., Makarov, V., Lakshmi, B., Findling, R.L., Sikich,L., Stromberg, T., Merriman, B., Gogtay, N., Butler, P., Eckstrand, K., Noory,L., Gochman, P., Long, R., Chen, Z., Davis, S., Baker, C., Eichler, E.E., Meltzer,P.S., Nelson, S.F., Singleton, A.B., Lee, M.K., Rapoport, J.L., King, M.C., Sebat,J., 2008. Rare structural variants disrupt multiple genes in neurodevelop-mental pathways in schizophrenia. Science 320 (5875), 539–543.

Williams, S.M., Sullivan, R.K., Scott, H.L., Finkelstein, D.I., Colditz, P.B.,Lingwood, B.E., Dodd, P.R., Pow, D.V., 2005. Glial glutamate transporterexpression patterns in brains from multiple mammalian species. Glia49 (4), 520–541.

Yang, W., Kilberg, M.S., 2002. Biosynthesis, intracellular targeting, anddegradation of the EAAC1 glutamate/aspartate transporter in C6 gliomacells. J. Biol. Chem. 277 (41), 38350–38357.